AACF
People & Skills ยท PPL-05

Role-based learning

People can get learning matched to what their role needs to do with AI, beyond the baseline for all staff, and use it in their work.

Why it matters

What a caseworker, a developer, a procurement officer and an auditor need to know about AI differs. Learning built around roles is more likely to change how people work than generic courses.

Levels
0 AbsentConfirmed absent: the evidence shows nothing is in place. Not known is recorded as insufficient evidence, not 0.
1 InitialLearning beyond the baseline depends on individuals finding their own courses.
2 DevelopingSome role-specific learning exists, but coverage is uneven and not linked to assessed needs.
3 DefinedResourced learning pathways, linked to the skills assessment, exist for every role group that works with AI, including suppliers, and cover risk duties and AI security.
4 ManagedUptake and the effect of learning on work are measured, and pathways with low uptake or effect are changed.
5 OptimisingPathways are updated from skills reassessment, feedback and changes in AI tools, and include learning on the job as well as formal courses.
Level 3 criteria

Level 3 is met when every applicable criterion is met. Criteria marked with an overlay apply only to organisations with that characteristic (overlays).

  1. PPL-05.1 Learning pathways are defined for each role group that works with AI and are linked to the skills assessment.
  2. PPL-05.2 The pathways are resourced and available across the organisation.
  3. PPL-05.3 The pathways cover the specific risk management duties of each role group.
  4. PPL-05.4 The pathways extend to suppliers and contractors who operate AI for the organisation.
  5. PPL-05.5 Pathways for roles that build, run, buy or secure AI include AI security.
  6. PPL-05.6 Pathways for people who build AI include secure design and development. BLD
  7. PPL-05.7 The AI security content of pathways is reviewed when significant new threats emerge.
Assessment questions
  • What learning is available for roles that build, buy, assure or use AI heavily?
  • How does it link to the skills gaps you have identified?
  • How do you know it has changed how people work?
Evidence an assessor would expect
  • Role-based learning pathways and the needs they address
  • Uptake records and evaluation results
  • Learning records for people with AI risk duties, including contractors
  • AI security in learning pathways for builders, operators, buyers and security staff
How to reach Level 3
  1. Identify the role groups that work with AI and, for each, write a learning pathway built from the gaps found in the skills assessment (PPL-04).
  2. Include in each pathway the risk management duties that role holds, and add AI security for roles that build, run, buy or secure AI, with secure design and development for builders.
  3. Fund the pathways, make them available across the organisation, and require suppliers and contractors who operate AI for you to follow the relevant pathway.
  4. Set a trigger to review security content when a significant new threat emerges, and name who carries out that review.

Common pitfalls

  • Offering a generic catalogue of courses that no one has linked to actual role needs.
  • Forgetting that buyers and risk staff need AI learning as much as developers do.
  • Assuming suppliers already know your risk duties and security rules.

Starter artefacts

  • Role-based learning pathway map
  • Learning records for people with AI risk duties
  • Security content review trigger log
Ownership and sequence
Accountable
people lead
Contributors
business owner, delivery lead, security lead
Depends on
PPL-01, PPL-04
Roadmap step
4
Lifecycle
design, acquire_build, deploy, operate
Cross-references
AI Cyber Security Code of Practice1.1 direct; 1.2 partial
Algorithmic Transparency Recording StandardTier 2 Deployment context partial
EU AI ActArt 4 partial; Art 26(2) partial
GDS AI PlaybookP9 direct
ISO/IEC 420017.2 direct
JSP 936 Part 1Preface para 7 partial; para 49c partial; para 61 partial; para 185 partial
MITRE AI Maturity ModelP3.3 direct
NIST AI 600-1R7 partial
NIST AI RMF 1.0GOVERN 2.2 direct; MAP 3.4 direct
UK Data and AI Ethics FrameworkSafety 9 partial
UNESCO AI Maturity Framework2.1 direct; 2.5 partial

direct: this control at Level 3 meets the item. partial: it covers part of it.